A procedure was developed to determine kinetic constants for competitive inhibition (Km, Ki, Vmax) by the median method. Simulated experiments were used to compare the accuracy and precision of kinetic constants determined by the median method with unweighted and weighted least-squares analysis. The median method was superior to unweighted least-squares analysis. The weighted least-squares method was superior to the median method when the error was normally distributed but the median method was superior when two or more outliers were present. The dependence of the accuracy and precision of kinetic constants obtained by the median method on several experimentally important parameters, including the number of experimental points, the number and range of substrate concentrations, and the number and range of inhibitor concentrations, was determined.